Speaker
Description
Robust control of the heat and particle exhaust is required for machine protection in future high power magnetic confinement devices such as ITER. Understanding the dynamic behaviour of the plasma boundary is essential both for control design [1] and more generally to improve knowledge of the relevant transient physics. Recently, the inherently dynamic MHD code JOREK [2] has been extended with a kinetic neutral model to improve its fidelity regarding the description of the scrape-off layer and divertor plasma [3]. This provides an opportunity to study time dependent heat and particle exhaust, for reasonable computational times at the ITER scale, and thus to construct a synthetic environment to explore divertor detachment control.
There is, however, still a gap between the kinetic neutral capability in JOREK compared to EIRENE [4], and the dynamics in the model have not yet been validated. To close part of this gap in the neutral capability, neutral self-collisions have been implemented in the form of a direct binary collision algorithm, which has been verified to reproduce diffusion correctly. This model is more accurate than the widely used BGK model (e.g. in EIRENE), which can only reproduce correctly one collisional effect at a time (e.g. viscosity or diffusivity).
To validate the dynamics in the model, dedicated experiments have been performed on ASDEX Upgrade in which a transient detachment state is initiated by a step in the gas fuelling rate. The discharge was an Ohmic deuterium L-mode (AUG shot #42651; P=0.7MW, Ip=0.8mA, Bt=-2.5T, $Γ_{D}=0.22 \mathrm{(initial)}-8.4 \mathrm{(final)}⋅10^{21} e/s$, κ=1.58, $δ_u$=0.33) with extensive diagnostic coverage to provide the best possible set of experimental measurements under the simplest conditions. After initializing the transient, the new equilibrium was reached within one second.
We present a comparison of simulations with the upgraded JOREK model including the impact of the neutral collisions against this experiment. Initial results indicate, as expected, low impact of the addition of the neutral collisions on the behaviour of the plasma in the initial attached phase of the shot. That the simulations are found to be more attached than in the experiment is also expected owing to the lack of molecular interactions in the model which are known to play a major role in detachment onset [5].
[1] Ravensbergen, Nat. Commun. 12(2021) 1105.
[2] Hoelzl, Nucl. Fusion 64(2024) 112016.
[3] Korving, Phys. Plasmas 30(2023) 042509.
[4] Reiter, Fusion Sci. Technol. 47(2)(2005) 172–186.
[5] Verhaegh, Nucl. Fusion 63(2023) 016014.